placed in adjacent holding pastures

with vegetation similar to that of the

sampled pastures. Animals were
allowed to graze untreated in holding
pastures for 10 days prior to sensory
Smell, Sight, and Touch impairment so they would become
familiar with the vegetation.
Two pastures were each grazed five
to Forage Selection successive days in four different trials.
Twenty sheep representing five treat-
ments were sampled in each trial. The
first trial compared forage preferences
of untreated fistulated sheep with
WILLIAM C. KRUEGER, WILLIAM A. LAYCOCK, AND DONALD A. PRICE those of sheep that had a single sense
impairment (Table 1). Trials 2-4 ex-
Highlight: Chemical impairment of taste, smell, and touch and physical amined forage preferences of sheep
obstruction of sight were studied in relation to forage preferences of sheep in a that had varying combinations of sen-
tall-forb plant community. Taste was the special sense most influential in direc- sory impairment and compared prefer-
ting forage preference; the other senses appeared to supplement taste. Sheep ences with those of untreated sheep in
preferred sour and sweet plants and generally rejected bitter plants, although each trial. Consequently, all possible
some were palatable. Smell was of minor importance in selection. Touch and combinations of treatments were
sight related to such specific plant conditions as succulence and growth form. included in the study.
Simultaneous impairment of all four senses did not result in completely random
selection, but did increase preference for unpalatable plants and decrease prefer-
Treatments to impair the senses of
ence for palatable ones.
smell, taste, and sight were performed
at least 4 days before sampling to
allow sheep time to adjust to treat-
Little is known of the effect of southwestern Montana. The study area ments. Suppression of the senses of
herbivores special senses in selection was 7,800 ft in elevation and domi- taste and smell (ageusia and anosmia)
of their diet; however, it is commonly nated by tall forbs. Eight l/2-acre was induced by topical application of
accepted that different plants within a study pastures were located on a 20% a 2% solution of osmic acid (0~04).
plant community have varying palata- slope that had a north aspect in a This chemical alters nerve endings,
sticky geranium/northwest cinquefoil making them insensitive to stimuli for
bilities for grazing sheep. Palatability (Geranium viscosissim urn/P0 ten tilla
of a plant species varies somewhat 6 months or longer (0~01 and Farrar,
gracilis) community. Two pastures
depending upon its association with 1955). Control of smell and taste was
were used in each of four grazing tested after treatments were applied
different plant species, but many fac- trials.
for several days and for several weeks.
tors interact to produce a selective Forage production of each pasture Untreated sheep reacted vigorously to
response. This selective response has was sampled immediately prior to each fumes from household ammonia;
been termed preference. An excel- grazing treatment. Except in one anosmic sheep reacted to ammonia as
lent review of important factors instance, 30 plots, each 4.8 ft2, were they did to water. Salt consumption
involved in forage selection by rumi- located by a stratified random method was measured for untreated and
nants has been presented by Marten in each pasture; prior to Trial 4, only ageusic sheep. When salt was juxta-
(1970). The study reported here was 15 plots were sampled. Major plant posed with pelleted alfalfa, ageusic
undertaken to gain insight into the species were clipped at ground level sheep consumed 8.7 times as much salt
basis of forage preference. and weighed, and green weight of as untreated sheep. When salt was
minor species was estimated. Samples physically displaced and sheep had to
Study Area and Methods of each plant species were collected make a deliberate selection for salt,
and ovendried at 50C to allow con- ageusic sheep consumed only half as
Field sampling was conducted dur- version of green weight to dry weight. much as untreated sheep. Sight was
ing the summer of 1968 on the U.S. controlled (anopsia) by means of a
Sheep Experiment Station summer Twenty-five esophageally fistulated
sheep, that had similar breeding and blindfold (Fig. 1). Sheep in the initial
range in the Centennial Mountains of anopsic group had blindfolds in combi-
experiences, were used in the study.
Twenty were yearlings; the rest were nation with other treatments through-
technician and out subsequent trials. Blindfold treat-
La y!$Zgeirs
l%Zt ZZ%gist , International
2-5 years old. Only 20 sheep were used
during a sampling period. However, ments varied for sheep that initially
Forest and Range Experiment Station, For-
est Service, U.S. Department of Agriculture, because of problems resulting from the had other treatments (Table 1). The
Ogden, Utah; both stationed in Logan, Utah, sense of touch in the lips was con-
at the Forestry Sciences Laboratory, main- fistulation, occasionally an unthrifty
tained in cooperation with Utah State Uni- yearling was replaced by one of the trolled (atactilia) by topical applica-
versity. Presently, Krueger is assistant profes-
older sheep. Sampling was begun at tion of 5% xylocaine ointment. This
sor, Rangeland Resources Program, Oregon
State University, Corvallis. Price is research dawn and fistulated sheep collected treatment was applied daily immedi-
leader, U.S. Sheep Experiment Station, Ag- samples for approximately 1 hour. ately prior to grazing. Xylocaine is a
riculture Research Service, U.S. Dep. Agr., topical anesthetic commonly used in
Dubois, Idaho. Prior to sampling, sheep were fasted
from dark until daybreak. During the medical practice.
The report is part of a dissertation sub-
mitted by the senior author in partial fulfill- sampling period, salt and water were Fistula samples were placed in a
ment of the PhD degree in range science at removed from the pastures so that 20-mesh sieve, washed in 5% acetic
Utah State University.
The authors acknowledge Dr. R. L. Hurst, chemical analyses could be performed acid to remove saliva, and rinsed with
head, Department of Applied Statistics and on the samples. Results of chemical water. The samples were then stored
Computer Science, Utah State University, analyses will be reported in another on dry ice and kept frozen until anal-
for assistance with statistical procedures.
Manuscript received October 20, 1973.
paper. Between trials, sheep were yzed.

258 JOURNAL OF RANGE MANAGEMENT 27(4), July 1974

 Plant species present in the diet
were determined by a modification of
Treatment groups
Heady and Torells (1959) technique.
Each sample was thawed, spread on a Trial Date (19681 1 2 3 4 5
tray, and 100 points were examined
through a stereoscopic microscope
I July 22-26 f-4 AS w Ag Ati
2 July 30.Aug. 3 N AP, AS AP> Ag As, Ag AP, At
that had a crosshair mounted in one 3 Aug. 7-11 N As, At AP, Ag, As As, At, Ag Ag, At
ocular. The plant fragment directly 4 Aug. 17-21 N As, Ap, At As,Ap,Ag,At - - - AP, At, Ag
under the crosshair was identified IN = control.
from epidermal characteristics. Plant
fragments that could not be identified
at 15X were removed, mounted on a
slide, and examined at 125X (Ward,
1970).
A relative preference index (RPI) a plant by a sheep, but it did prove is presented in Table 2. Sticky gera-
was used to evaluate forage prefer- useful in interpreting the grazing pre- nium made up about 50% of green
erences and to minimize confounding ference of the animals. production on a surface area basis and
differences in preference that might Statistical comparisons were made had a frequency ear 100%. Northwest
result from varying availability of by sing least-squares techniques cinquefoil, second in availability, made
plant species (Krueger, 1972). The (Draper and Smith, 1966) and planned
up about 10% of the composition and
following formula was used to deter- contrasts (Ostle, 1963). Each treat-
mine relative preference for each plant had a frequency greater than 90%.
ment was compared to the control for
species in each diet. Values were based the same sampling period; no statis- Mountain brome (Bromus merginatus)
on surface area of _ green .lants: tical comparisons were made across made up about 8% of the composition
fd X D periods or among other treatments. and had a frequency about 90%. The
Relative Preference Index = __
fr X R Differences were accepted as signifi- most palatable plants in this commun-
Where: fd = frequency in diet cant at P < 0.20. At P < 0.20 results ity, mountain sorrel (Rumex
D = percent diet of statistical testing were consistent paucifolius) and thickleaf grounds4
fr = frequency on rangeland and biologically explainable. At lower (Senecio cr~ssulus), made up about 1%
R = percent range composition levels inconsistencies not biologically
of the composition and were lower in
explainable began to appear in data
Frequency in diet was the percent- frequency than the more abundant
analysis.
age of samples for each treatment in species.
which a species occurred. Percent diet Relative preference exhibited by
was the percentage of a plant species Results
untreated sheep varied for each time
in each diet on a surface-area basis All pastures sampled during the period. The effects of each treatment
determined from microscopic point study were similar in amount of each on preference were evaluated by com-
sampling. Frequency on rangeland was plant species available to the grazing paring the control to the treatments
the percentage of plots in which a
sheep. To help develop a perspective studied only during a common trial.
plant species occurred for each pasture
of the plant community structure. The average of each treatment was
grazed during each trial. Percent range
composition was determined from average composition on a green surface computed to determine the relative
values for green weight converted to area basis for important plant species preference of sheep for the plant
surface area by using a small stage
photoelectric planimeter (Caldwell and
Moore, 1971). Interception of trans-
mitted light for known green weights
of samples of each plant species was
compared to a standard curve (r =
0.99) and surface area in relation to
weight was determined. All green
weight production figures were then
converted to percent surface area for
use in the preference index.
A subjective evaluation of coarse-
ness, smell, and taste was based on
human sensory perception. Coarseness
was evaluated by chewing the different
plants and assigning them to one of
three categories (high, moderate,
or low coarseness). It was difficult
to assign a specific smell to each plant
species, although there were many
distinct odors among the various
plants. Smell was classified as
strong, moderate, or faint.
Taste was classified as bitter,
sweet, or sour. The categorization
of these plant attributes may or may
not be equivalent to the evaluation of
species studied in these pastures (Table Table 2. Range composition (%) of major plant species based on green surface area, and
3). Buchanan et al. (1972) have pre- total production (lb/acre, ovendry) by trials.
sented preference rankings for these
Measurement and
plants from fistulated, but otherwise Trial 3 Trial 4
plant species Trial 1 Trial 2
untreated, animals grazing the same
range. Similar palatability relationships Composition
were found in their study when these Grasses
were compared to forage preferences Alpine timothy 3.8 2.4 1.2 1.2
Mountain brome 8.1 7.9 9.9 8.5
of control sheep in this study. Showy onion grass 2.4 .7 1.4 1.8
Although all senses were related to Forbs
Duncecap larkspur 3.3 2.9 1.9 2.8
preference for some plant species, Leafybract aster 5.4 3.6 4.6
5.5
taste appeared to be the most impor- Mountain knotweed 1.2 .8 .4 2.4
tant. Sticky geranium, classified as Mountain sorrel .8 .4 .2 .2
bitter , was unpalatable. Compared Northwest cinquefoil 10.6 11.7 9.2 8.0
Pale agoseris 1.5 1.0 .4 .2
to untreated sheep, ageusic sheep sig- Stickseed 2.5 1.4 1.4 1.0
nificantly preferred sticky geranium Sticky geranium 39.8 48.8 57.4 49.0
(Table 4). When all senses were im- Sweetanise .6 .9 1.9 1.0
paired except taste, preference did not Thickleaf groundsel .6 1.4 1.2 1.0
Western valarian 3.7 .8 .7 .7
deviate significantly from that of con- .9
Western yarrow 1.5 .6 1.2
trol animals. When either smell, sight,
Total production 1,827 1,748 1,516 1,570
or touch was controlled, preference of
sheep for geranium was not sig-
nificantly different from that of pleasant taste and smell, bitter-tasting showed a preference for larkspur of
untreated animals. However, when western yarrow was also recognized by 5.6 compared to 0.8 for control sheep
taste was impaired in combination sight as being unpalatable. in the same trial. When sheep could
with one, two, or three other senses, Both duncecap larkspur (Delphinium see, smell, or taste, regardless of com-
the sheep showed a significant increase occidentale) and northwest cinquefoil, bination of treatments, preference for
in their acceptance of sticky geranium. both bitter-tasting plants, were more larkspur appeared to be lower than
Palatability of western yarrow palatable when sheep had their sense when these senses were controlled.
(Achilles lanulosa) was also related to of taste controlled. Because of the low Frequently, however, no statistical dif-
taste, but not as exclusively as that of availability of larkspur, valid statistical ference was found.
sticky geranium. The only sense not comparisons could not be made for all Northwest cinquefoil was another
eliciting rejection for yarrow was trials. Duncecap larkspur seemed to plant that significantly increased in
touch, an indication that western yar- have an objectionable smell and taste. preference when taste was controlled,
row possesses many attributes that Touch was not related to preference but its palatability was not exclusively
lower its palatability. In addition to for larkspur, but sight appeared to be related to taste. Impairment of smell
having what seemed to be an un- negatively related. Blindfolded sheep decreased preference for northwest

Table 3. Average relative preference index for treated and untreated fistulated sheep in four trials.

Grasses Forbs
Dunce- Moun- North- Thick-
Moun- Showy cap Leafy- tain Moun- west Pale Sticky leaf
Alvine tain onion- lark- bract knot- tain cinque- agos- Stick- gera- Sweet ground- Western Western
Trials timbthy brome grass - svur aster weed sorrel foil eris seed nium anise se1 valarian varrow
No. 1
Control 1.2 2.0 1.5 .8 5.0 .6 11.9 1.3 4.0 1.0 LZ 36.4 55.7 1.7 L
AP 1.8 2.4 2.0 5.6 5.3 .9 7.9 1.4 3.0 .5 L 7.8 45.8 1.6 .3
As 1.4 1.5 1.7 7.7 5.2 .9 11.1 1.0 4.2 1.4 .2 19.5 73.1 3.9 .3
Ag .2 .8 1.1 9.5 4.7 1.0 2.5 1.7 5.8 2.3 .2 24.8 31.5 1.4 .2
At 1.4 2.4 1.9 1.2 4.6 .3 .9 1.4 2.2 1.6 .l 9.9 31.9 1.1 .l
No. 2
Control 2.2 2.2 7.6 .5 6.3 1.7 37.4 1.1 3.4 .9 L 4.6 14.2 1.3 .2
AP, As 2.4 2.2 13.8 L 5.6 2.6 29.6 1.1 1.4 .8 .l 8.3 10.0 3.5 .4
AP, Ag 2.5 2.2 9.2 .4 5.4 1.4 .2 1.5 1.8 .5 .l 5.5 6.6 1.0 .8
As, Ag 1.2 1.4 7.8 L 5.0 1.0 2.2 2.0 2.8 2.5 .l 5.9 5.6 3.7 .5
AP, At 2.3 2.1 9.6 L 4.5 .3 11.4 1.1 .9 1.0 .l 17.8 2.9 .5 .2
No. 3
Control 18.5 2.1 3.7 L 9.6 3.7 69.6 2.0 25.6 .7 L 14.6 12.2 2.0 .l
As, At 1.2 1.8 3.0 L 7.8 8.2 61.4 1.4 3.2 4.5 .l 8.2 13.9 .l .5
As, AP, Ag 7.0 2.2 4.9 35.1 7.5 4.8 2.4 1.8 13.2 .5 .2 4.1 20.2 2.1 1.4
As, Ag, At 5.8 1.2 3.8 1.9 9.4 2.4 9.1 2.1 3.5 3.5 .2 3.3 10.6 1.8 1.0
Ag, At 4.7 1.8 4.4 L 9.6 4.3 5.3 1.4 6.8 1.9 .l 2.7 27.8 8.0 .3
No. 4
Control 2.0 2.6 2.2 .3 12.8 .8 2.4 2.1 2.4 1.9 L 2.5 7.2 .4
As, Ap, At 1.2 2.9 2.8 .4 11.9 1.3 1.8 1.2 1.9 .4 .l 2.4 13.2 :: .2
AP, At, Ag 2.4 2.6 3.4 .l 5.7 .4 .5 1.7 .l .4 .l 4.0 .l 2.1
As, Ap, Ag, At 2.4 2.8 3.1 .3 8.0 1.6 .l 1.0 2.7 :: .3 2.7 3.2 1.4 2.1
Symbols used are as follows: As = anosmia, sense of smell impaired; Ap = anopsia, sense of sight impaired; Ag = ageusia, sense of taste impaired; and At = atactilia, sense of
touch impaired.
L = less than .l.

260 JOURNAL OF RANGE MANAGEMENT 27(4), July 1974

Table 4. Statistical relationships of relative preference indices for selected treatments of sheep when compared to control sheep. Values indicate the
level of significance of difference when compared to control.

As, AP,
I
As, Ap, At As AP, At, Ag AP As, Ag, At As, AP, Ag Ag, At
Plant species $2 (4) (1) (4) (1) (3) (3) (4)
Grasses
Alpine timothy .02 NS3 NS NS .20 .15 NS .20 NS
Mountain brome .Ol .Ol NS .Ol NS NS NS NS .02
Showy oniongrass NS NS NS NS .20 NS NS NS NS
Forbs
Duncecap larkspur .05 NS .lO NS NS NA4 NS NA NS
Leafybract aster NS NS NS .02 NS NS NS NS NS
Mountain knotweed NS NS NS .20 NS NS .20 NS .05
Mountain sorrel .Ol NS NS .20 NS .lO .Ol .lO .lO
Northwest cinquefoil .lO .05 .15 .20 NS NS NS NS .lO
Pale agoseris NS NS NS .15 NS .lO .lO NS NS
Stickseed NS .05 NS .05 NS NS NS NS .05
Sticky geranium .Ol NS NS .Ol NS .Ol NS .Ol .Ol
Sweetanise NS NS NS NS .20 .02 .20 .05 NS
Thickleaf groundsel NS NS NS .15 NS NS .20 NS .20
Western valarian NS NS .lO NS NS NS NS NS NS
Western yarrow .02 .15 .Ol .20 .Ol .15 NS .02 .15
Symbols used are as follows: As = anosmia. sense of smell impaired; Ap = anopsia, sense of sight impaired; Ag = ageusia, sense of taste impaired; and
Ai = atactilia, sense of touch impaired.
2Trial number.
Not significant at P < .20.
4Not statistically tested.

cinquefoil; anosmic sheep had a prefer-
ence value of 1 .O compared to 1.3 for
control animals. However, the interac-
tion of smell and taste suggested that
taste was more influential in selection
of this plant. When both smell and
taste were controlled, the relative pre-
ference index increased, which is the
same response observed with impair-
ment of taste alone, but opposite of
the response observed when smell
alone was impaired (Table 3). Although
this difference was not significant, it
did show a tendency for the effects of
taste impairment to override the
effects of smell impairment. Further
evidence to suggest a more important
influence of taste was found when all
other senses were impaired. When this
occurred, preference decreased signifi-
cantly; animals that could only taste
took less northwest cinquefoil than
control animals. When senses except
smell were controlled, sheep prefer-
ence remained depressed, which indi-
cated that selection on the basis of
smell was dependent upon other sen-
sory stimuli.
Preference for grasses was primarily
related to taste and sight. Preference
for alpine timothy (Phleum alpinum)
and for mountain brome was signifi-
cantly decreased when taste was
impaired. Preference for showy onion-
grass (Melica spectabilis) was not sig-
nificantly altered by taste treatment.
Preference for showy oniongrass and
alpine timothy increased when sight
was impaired. This change was prob-
ably related primarily to the change in
feeding behavior of the sheep when
blindfolded. Anopsic sheep tended to
feed longer in a specific spot and lower
in the vegetation. Grasses that tended
to grow lower than most other vegeta-
tion were more frequently en-
countered by anopsic sheep than by
untreated animals. Since these low-
growing grasses did not have other
o b j e c tionable characteristics, they
were consumed more frequently by
anopsic sheep than by sheep that had
better access to higher growing vegeta-
tion because of normal behavior.
Changes in feeding behavior were only
apparent for anopsic sheep.
Mountain sorrel was a forb strongly
preferred because of its sour taste.
When taste was impaired singly or in
combination with other senses, prefer-
ence for sorrel decreased. Mountain
sorrel a tender, succulent plant, was
also preferred on the basis of touch.
However, taste seemed to be more
influential, as indicated by results
when all senses except touch were
controlled. In this case, the preference
index for treated sheep was 2.4 com-
pared to 69.6 for untreated sheep.
Preference remained depressed when
all senses except touch were impaired.
When all senses except taste were
controlled, no significant differences
were noted between treated and con-
trol sheep.
Preference for other succulent
plants, such as thickleaf groundsel,
swee tanise (Osmorhiza occiden talis),
and pale agoseris (Agoseris glauca),
was also positively related to the sense
of touch. Mountain knotweed (Poly-
gonum montanum), not succulent but
rated low in degree of coarseness,
decreased in preference when touch
was controlled. No plants had charac-
teristics producing sufficiently adverse
tactile stimuli to cause significant
decreases in preference. The role of
touch in selection was not as influen-
tial as that of other senses. Preference
for thickleaf groundsel was more re-
lated to interactions of smell and taste.
Sheep with smell and taste controlled
showed a significantly lower prefer-
ence (5.6) than untreated sheep (14.2)
with P < 0.01. Most bitter plants in
this study were unpalatable; however,
the bitter-tasting thickleaf groundsel
was preferred when smell and taste
functioned together. Since preference
for groundsel decreased when only
smell remained intact, it is probable
that the bitter taste of this plant and
its tenderness provide highly preferred
stimuli for acceptance by sheep.
Sweetanise was a highly preferred
plant. Acceptance of this plant
decreased from RPI = 36.4 to RPI =
24.8 when taste was controlled. While
the difference was not significant, it
did suggest sheep preferred the sweet
taste of this plant. Sweetanise, also
highly succulent, was preferred on the
basis of touch. Sight was also related

JOURNAL OF RANGE MANAGEMENT 27(d), July 1974 261

to palatability of sweetanise, which is
a tall-growing plant. Sheep could see
sweetanise plants from any area of the
pasture and would actively seek them
out unless blindfolded. Other plants of
exceptionally high palatability were
not as obvious to sheep at long dis-
tances, and the animals did not signifi-
cantly relate to sight of them.
Leafybract aster (Aster foliaceus)
was both abundant and high in palata-
bility. This aster has a faint bitter
taste, but is still highly preferred by
sheep on this range. Changes in prefer-
ence for aster as a result of sensory
impairment generally were not signifi-
cant. Smell apparently was not of
importance in selection of aster. Pre-
ference decreased significantly when
smell was the only sense intact, indi-
cating that other senses related to
preference but smell was neutral. It
was suspected from results of treat-
ment combinations that touch was
influential in determining preference
for leafybract aster, but no clearcut
significant relationship could be estab-
lished.
Stickseed (Hackelia j7oribunda) was
selected nearly at random by the
grazing sheep. No relationship to spe-
cific senses could be established,
although there were indications that
sense alteration would change prefer-
ence values.
Discussion
Results of this study indicated rela-
tionships of the senses to forage selec-
tion are complex. Taste appeared to be
the primary sense used for selection of
many species in the animals diet,
although taste frequently interacted
with other senses. However, taste of
some plants had no apparent influence
on their palatability. Arnold (1966b)
found that sheep surgically treated to
control the sense of taste were unable
to differentiate between species and
strains of plants as successfully
intact sheep. This work was done on
improved pastures in Australia, but the
basic relationships of taste seem to be
the same as on rangeland in this study.
Longhurst and Kepner (1968) studied
the effect of impairment of smell and
taste in selection of feed treated with
extracts from plant leaves high in
volatile oils. They found taste to be of
less importance in preference than
smell. However, when smell and taste
were both controlled, the degree of
selectivity was less than if smell alone
was controlled. palatability, such as mountain sorrel
On the range studied, sticky gera- and thickleaf groundsel, decreased in
nium made up about half of dry preference. Although sensory impair-
weight production. Since selectivity of ment was severe, some factors influ-
this plant was almost exclusively a encing forage selection, such as touch
function of taste, some change in the in the tongue and the common chemi-
taste of this plant to improve palata- cal sense, were not controlled. Arnold
bility could almost double the carrying (1966b) reported the control of smell,
capacity. Perhaps a chemical substance taste, and touch in combination did
could be applied to improve the taste not completely alter animal forage
of geranium or a fertilizer program preference in an area where sight was
might change palatability of this plant. of minor importance. Longhurst and
The other senses appear to supple- Kepner (1968) in feeding trials of
ment the sense of taste. Smell was sheep also found impairment of smell,
only of specific importance in selec- taste, and touch in combination did
tion of northwest cinquefoil; in con- not completely alter sheep forage pre-
ferences.
junction with taste, smell was influen-
tial in selection of certain other plant
Literature Cited
species. Tribe (1949) suggested smell
&mold, G. W. 1966a. The special senses in
was only supplementary to other grazing animals. I. Sight and dietary
senses in influencing forage selection habits in sheep. Aust. J. Agr. Res.
of sheep. Arnold (1966b) found smell 17:521-529.
to be of importance and closely Arnold, G. W. 1966b. The special senses in
related to selection of plant parts or
grazing animals. II. Smell, taste, and
touch and dietary habits in sheep. Aust.
specific phenological stages of plants. J. Agr. Res. 17:531-542.
Longhurst and Kepner (1968) deter- Buchanan, H., W. A. Laycock, and D. A.
mined that olfaction was the first Price. 1972. Botanical and nutritive con-
sense used in making a selection and so tent of the summer diet of sheep on a tall
had its greatest effect in initial forage
forb range in southwestern Montana. J.
Anim. Sci. 35:423-430.
selection. Caldwell, M. M., and R. T. Moore. 1971. A
Sight was related to selection of portable smallstage photoelectric
certain palatable plants, such as planimeter for leaf area measurements. J,
sweetanise. Other plants of high pala-
Range Manage. 24:394-395.
Draper, N. R., and H. Smith. 1966. Applied
tability, but low-growth form, could regression analysis. John Wiley and Sons,
not be seen from distances greater Inc., New York. 407 p.
than 10 or more feet. Consequently, Heady, H. F., and D. T. Torell. 1959.
no relationship to sight was shown. Forage preferences exhibited by sheep
with esophageal fistulas. J. Range
Arnold (1966a) found sight was not Manage. 12:28-33.
important for selection of plant Krueger, W. C. 1972. Evaluating animal
species in grass-clover-alfalfa pastures, forage preference. J. Range Manage.
but did report changes in grazing 25:471475.
Longhurst, W. M., and R. E. Kepner. 1968.
behavior similar to those described in
Strontium-90 and Cesium-137 food chain
this paper. Arnold did not blindfold relationships to deer and sheep. V. Fac-
the sheep; he restricted their vision tors influencing radionuclide accumula-
with blinkers that did not totally tion. Progress Report. AEC Project No.
obstruct sight except at the point of UCD 34, p. 104-112. (68 p. mimeo.)
Marten, G. C. 1970. Measurement and sig-
forage selection. nificance of palatability. In: Proc. of
Plants selected, at least in part, on National Conference on Forage Quality
the basis of touch were all categorized Evaluation and Utilization. Dl-D55.
as as low in coarseness; generally they Nebraska Center for Continuing Educa-
tion, Lincoln.
were succulent. Overall, touch
0~01, A., and G. E. Farrar, Jr. 1955. The
appeared primarily to supplement the dispensatory of the United States of
sense of taste in selection of plant America. J. B. Lippincott Co., Philadel-
species. phia, Pa. 2139 p.
Ostle, B. 1963. Statistics in research, 2nd
When all four senses were impaired,
ed. Iowa State Univ. Press, Ames. 585 p.
sheep showed some interesting altera- Tribe, D. E. 1949. The importance of the
tions in forage preference. Changes in sense of smell to the grazing sheep. J.
preference did not result in completely Agr. Sci. 39:309-312.
random forage selection; however, Ward, A. L. 1970. Stomach content and
fecal analysis: Methods of forage identifi-
plants of low palatability, such as cation. In: Range and wildlife habitat
geranium and western yarrow, in- evaluation-a research symposium, p.
creased in preference after impairment 146-158. U.S. Dep. Agr. Misc. Publ.
of all four senses. Plants of high 1147.220 p.

262 JOURNAL OF RANGE MANAGEMENT 27(4), July 1974